Reboot unknown

[Russ Walicki]

Anyone,

I have a B132L+ HP10.20 workstation and for some reason today it has been rebooting "on it's own" many times. Sometime it will run for hours other times it will run for minutes and currently it will only run about 30 minutes before it reboots for no known reason. syslog.log shows nothing out of the ordinary?

Does anyone out there have any ideas on how to determine the root cause?

Moreyrd

[Rajeev Shukla]

It looks like there is some hardware problem with the PowerSupply of the worrkstation. Also check for any fluctuation of power in the server room. Could be ventilation problem if the server is getting heated too much.
All problems related to power etc will not be reported in the syslog.

Cheers
Rajeev

[Jeff Schussele]

Hi Robert,

Couple of things:

1) Check the /etc/shutdownlog file for an indication of what started the shutdown.

2) Check the /var/adm/crash dir for tombstone file(s) - would be named tsxx where xx=99 would be the most recent.

3) Determine where your dump files reside (usually /var/adm/crash/core*) and check the INDEX file for a panic statement. You may need to further analyze the dump file with q4 commands to see what really happened.

HTH,
Jeff

[Michael Tully]

You'll probably find you have a faulty piece of equipment, CPU, backplane. First see if your system is actually panicing by having a look at the /etc/shutdownlog. Then have a look at the tombstones directory for the file differences. /var/tombstones.

[Bill Hassell]

Your system is crashing (the Unix term is panic), probably due to lack of patches. When Unix crashes, there will be nothing recorded in syslog because Unix has halted. You probably don't have an /etc/shutdownlog so create it. Now in order to see the reason for the crash, you'll need to create /var/adm/crash so a crash dump can be created. However, this directory must be at least as large as RAM.

Once /var/adm/crash is available abd /etc/shutdownlog exists, you'll see a simple record of the reason for the crash in /etc/shutdownlog. If it says HPMC, then you have a hardware failure, probably CPU or power supply. Otherwise, it is likely a lack of patches.

[Heinz Gaertner]

Hi,

got this on the same machine long time ago. There it was the firmware-release (the PDC) which was the problem. If i remember you can see the actual version in the boot menu. We had downloaded a newer version and everything was fine. Could be worth a try.

Heinz

[Bill McNAMARA_1]

Are you running any odd applications at the time? How well are you patched?
Sometimes (apart from HW/cooling failures) it's not worth much to look too closely at the PIM and HPMC unless you're well patched.

Later,
Bill

[Russ Walicki]

Thank you to all. We have determined the problem was related to power fluctuations due to a bad UPS battery.